1398 posts • joined Friday 15th June 2007 09:17 GMT
Before people start, the term "userspace" used in the PDF does not mean from a non-privileged process. It needs to be run as root or another ID with write permission to /dev/mem.
What "userspace" means here is a process run as a normal process controlled by the scheduler, and not added from inside the kernel codebase (like a loadable kernel module would).
Basically, all this technique is doing is re-vectoring one of the system calls, something that people have been doing for as long as table driven vector entry for system calls has existed. UNIX has done things this way since it first existed 40+ years ago (it was very convienient in the PDP/11 world, as it used the EMT instruction). The only real trick here is reserving memory in the kernel address space, and even this is not new (I could probably think of about hald-a-dozen candidates for locating the code off the top of my head).
Due to a design flaw in the MT10 magtape driver code in Bell Labs UNIX version 7 for the PDP/11 (circa 1978), we used to hang the tape device moderatly frequently. I used to go in and zap the lock bit in the driver status table using db (the original UNIX debugger) to use it again without re-booting. And the Keele Kernel Overlay system used to re-vector all of the system calls to allow segmentation registers 6 to be altered to point to the area of memory that had the required code, before actually jumping to it. This was all done in kernel space, of course, but show that the techniques are not new.
So. Stop frightening the ordinary users with things most of them will not understand, and just say that if you allow root access on your Linux box to any-old-code, expect your system to be 'pwned'
Most major distros actually ship with SEL turned off.
There are not that many applications that would break if it were turned on, but the administration of the Linux system would need to be changed. As a UNIX luddite, (and by this I mean someone who has been using it for so long that fundamental change appears abhorent), I can understand this, and I real uneasy about turning SEL on on my own systems. I am keenly aware that the UNIX security model, which Linux (pre SEL) copies almost exactly, has always been weaker than it could have been (although much better than Windows up to Vista). The MULTICS model that VMS and PRIMOS implemented would have been better from the start, but UNIX was intended to be lightweight compared to MULTICS.
But, as the major variant of UNIX that I use in my professional life is implementing Role-Based-Access as well, I guess that I will just have to learn to live with it.
The MAC attacks are DOS attacks, and reading through the PDF on the Linux attack, firstly is it x86 specific, and secondly, to exploit it you need WRITE access to /dev/mem or /dev/kmem (it's slipped in to the end of section 3 that this is required, and the test here is being run from a # prompt, indicating root access).
*NIX security 101 states that these should be protected from write (and even read in many cases), for just this reason.
Of course, if your vector runs as root, then all bets are off, and there are innumerable ways of making a *NIX system do bad things, even if you have SE turned on.
Same for DVB for laptops
I had a similar conversation a few years ago when I was inquiring about a DVB tuner on a laptop. It is illegal if it is plugged in to the mains outside of the house that the license is for, unless that location has a license that covers it.
One of the most stupid pieces of legislation I have ever heard.
Also, if you have a shared occupancy dwelling, like a student house, you need a separate license for each room where one of the occupants has a television, plus one for the communal area. Talk about greedy. You can only have a multiple televisions in a house on one license if everyone is a member of the same family. So, if you have a lodger in your house, they need a license as well!
I'm not actually against the license, as it allows the BBC to do things that otherwise would not be done (I cannot fathom a situation where Rupert Murdoch controls what television programs are made), but the heavy handed and way it is applied gets my goat.
Also, in every mail, they asked me what the number of the television license was for my house. They would not take my word that I actually had a license.
I remember listening to a House of Lords debate on the ID card. There were strong feeling from the opposition and the cross-party benches that the ID database would suffer scope creep, because it allowed the legislation to extend the database to be passed by statutory instrument rather than having it debated in parliament.
This has always worried me, and now it seams to have been worth worrying about.
I endorse a comment above. I don't mind carrying ID if it actually helps me prove my identity. I just don't want it to be used to track my activities. Not that I do anything illegal, but....
Try using your Epson printer on Windows *WITHOUT* using the supplied CD or downloaded driver from Epson. I think you would probably have more problems than on any mainstream Linux.
Badger the printer providers to give you either an install CD for Linux, or instructions to configure the excellent built-in Linux printer subsystem (like CUPS) to work with a new printer.
Don't use different rules to compare Linux with Windows.
Mind you, I do agree on other peoples comments about Xandros. My (very early) Eeepc 701 is running Ubuntu Hardy (I've standardised on an LTS release), because I got tired of re-installing Xandros each time the UnionFS filled up. Don't think it's a generic UnionFS problem, more a problem with the way it was configured.
When I read the title, I thought that amanfromMars had joined ElReg as a writer!
There is a blame culture in the UK. As a result, there is continued effort to eliminate human error in many 'systems' (I'm using system in the generic sense). Too often, this is done by taking the decision making away from real people, and codifying it according to unvarying rules.
This is absolutely fine, but only as long as you catch EVERY possible situation that the system has to cover.
But when you get a situation that you don't cover, chances are that you will get an inappropriate result. In the meantime, your human people, who have become de-skilled (either by accident or design) because they don't have to make these decisions, are less likely to spot the inappropriate response (they keep being told that the 'system' will do the checking, so they don't have to). So they blame 'the system', and are in many cases correct in doing so.
The failure is in the creation of the rules in that poor requirements and systems analysis has been performed. This makes the 'system' flawed, but as a result of a human failing (it could be a systemic failure in the process that created the system in question - recursion here we come!)
Please note that this is not limited to computer 'systems' but can happen to any process. It just so happens that so many complex systems nowadays are centred on computers enforcing the rules.
Only the English...
... would take a fuel that works well in an efficient internal combustion engine, and use it to generate steam to propel a car.
Now if they could achieve that speed using heavy fuel oil, or even better, powdered coal, then it may have some merit!
As far as I remember (and it gets more difficult as time goes by), student bars still buy their beer from brewers, at whatever price the brewers charge. There is no difference in duty charged, and many colleges use profits from the bar to generate money that gets spent on other subsidised activities.
About the only scope I can see for beer being cheap is that they probably do not pay rent, rates or energy costs (these will be soaked up by the college or Union), and that the staff work for peanuts, or at least beer money.
So, as pointed out, where is there scope for significant price reductions over places like Wetherspoons. Setting a price above what is paid outside the college will just reduce the use of the college bars, at least in colleges in large towns and cities.
What happens in out-of-town colleges like Lampeter or Keele is a different matter, and one that it may not be safe to pry into!
@Adrian Waterworth re B5
Where B5 actually was better was that the story ARC was fairly well populated for all 5 series (OK, it got a bit messy when Time/Warner canned it at the end of Series 4, and the order of the stories was ripped apart).
This meant that meaningful hints could be seeded throughout the series, which when spotted led to a big Ahhhhhh moment that spanned the seasons. This is what made B5 seasons 2-4 compulsive viewing for those of us who actually followed the storyline. But each episode stood-alone enough to be watched in isolation. I admit B5 was a mixture of absolute brilliance (like the episode that was filmed from start to finish in one scene - "Insurrections in real time"), and complete pap (I tend to forget these!). But you must remember when calling it cheesy, that for the first series at least, the video effects were produced on "the video toaster", which was a network of Amiga PC's.
I lost track of BSG when Sky 1 dropped off of Virgin's cable service, and found that whenever I did get to see an episode, that what I was watching made very little sense. Partly my fault, I know, but it did not make it easy to watch. But it seams to me that each series had it's own ARC, and not a lot was carried across from one series to the next.
Now, from B5, all I have to say is... "Get the hell out of my Galaxy!"
Point missed about MS contributed tests
Microsoft are obviously contributing tests to W3C that IE8 pass, but that Opera and Firefox fail.
It's probably the only way they can drastically improve their score compared to the others!
Does anybody remember that the term 'chicklet keyboard' was originally coined for the granddaddy of all PC's (possibly the original non-hobbyist microcomputer system), the Commodore Pet 2001?
Still happy with my EeePC 701. Must check out Eeebuntu though.
@AC re TDMA again
I follow all of your points, and I guess that I was simplifying things a bit, but it sound like a lot of effort, and synchronisation between different parts of the organisation, including putting engineers on the street physically fiddling with the cables. Must be some form of cost/benefit analysis on the value of this.
I'm not sure about the time-slot allocation technique for the bandpass filter. I'm fairly certain that they will not be "effectively connecting and disconnecting according to the time-slot allocation", but will have packet selection based on MAC, rather than an electronic time based switch. And it also does rely on being able to isolate the modem being investigated to a fixed carrier, preferably not shared with other modems.
I hate to think what the jitter of several dozen modems switching their bandpass filters on an off on a particular branch would be. Probably almost impossible to analyse.
Anyway, interesting discussion, as always on El. Reg.
No common Hardware model for Sun systems?
What is it about the T1, T2 and Niagara that make porting Linux so hard? Do they not run Sparc code from older systems? Or is it that the memory model or I/O subsystems of the resultant systems do not match what has gone before.
I admit that I am not following Sparc developments that closely, because I work mainly on IBM systems, but as far as I am aware, and as a comparison, there is no real need to port the Power based Linux distributions onto Power6, the current ones just work. Of course, there is a quite high degree of commonality between Power5 and Power6 systems (and to older systems as well, as chrp is a a common model), and you have all of the built-in virtualization which may help to isolate the OS from the hardware. And you probably won't get the full benefit of new features, but it works.
Have SUN not got that degree of abstraction from the hardware?
@AC re TDMA
I'm not a cable engineer, but am interested in the subject.
I'm not sure that it is that simple. The cable network, even from the street boxes is a tree network, with many branches and bi-directional signal regenerators. The signal is not baseband, but true multi-frequency broadband (my, how that term is mis-used in the public space), with many customers appearing on the same branch of the tree. It is not a point-to-point network like the telephone network.
The modulation is a mixture of TDMA and FDMA, with modems switching frequency during normal operation, and possibly using more than one frequency carrier for the higher data speeds.
If you were to inject noise (that is, without disconnecting the individual tap from the network), you would take out ALL customers ON THAT BRANCH! If you wanted to try to make it more selective, you could try to identify the frequency currently in use, but you would still take out all the customers using that carrier on that branch, and that is assuming that the modems would not switch to another frequency. If you want to check each tap, you would have to physically visit each tap point. And you must remember that this is a shared infrastructure with their cable TV operation.
You may also get false positives. What happens if, at the same time as you trying to identify an illegal modem, a customer turns off their cable modem?
I'm not sure a TDR (if you mean Time-Domain Reflectometer) would help either. These are used to measure cable breaks by relying on reflected signals from the end of cables without load (un-terminated in the transmission line vernacular). In this case, the cable would not be un-terminated, but would still have a load on it.
In theory, if you knew the prorogation time of the signal on the wire, it may be possible to time a response to the modem, but I suspect that the quality of cable, number of taps, and even the moisture content of the soil around the cable may alter the inductance and capacitance of the cables to make this uncertain. We're not talking 10base5 Ethernet, or even telephone line twisted pairs here.
All in all, I expect that the cable engineers at Virgin Media, who actually maintain a cable network, to know more about the design and running of these things than a majority of us amateurs commenting in this thread. Give them some credit, because if they knew nothing, their whole network would grind to a halt very quickly.
I've now tried to set up Wireless on XP (don't know Vista, no space for it in my house) using Microsoft Zero Config twice, and each time I've given up, because I just could not get it to work. As soon as I used the vendor supplied tool, it worked fine. I could be missing something here, but Zero Config is a lemon as far as I can tell.
Of course, the Belkin, Linksys, and Dlink wireless apps are all different, so you cannot really say that it is Windows that makes it easy if you use their apps.
For Linux, the secret is to choose a wireless chipset that is well supported. Once you do this, Network Manager (for all it's other faults) works fine. Of course, trying to work out what chipset is used is another problem entirely.
I implore all hardware vendors to support Linux as well as they do Windows. Then people would not blame 'Linux' for it's hardware support.
@avi re Sparc Thinkpad
Sorry. In case you hadn't noticed, IBM no longer own the Thinkpad brand, and the Lenovo/IBM joint marketing is in the wind-down phase, so no chance of a Sparc Thinkpad. I'm sure that there were Sparc laptops before (Tadpole? Solbourne?)
Not sure it would be work, any way. IBM had PowerThinkpad (830 and 850 models) with PowerPC 603e processors running AIX 4.1.3 (and, I believe WinNT on Power and even a Power OS/2 port) about 15 years ago, and decided there was no market for them. Mind you, they were real paving slabs (they were bigger than bricks) and very heavy.
Considering how much nostalgia gets written in these comments, isn't it about time there was a rose-tinted-glasses icon? That's what is in the top pocket of my coat, where I can find them easily.
Is Mike Mercury around to drive it? And are the boffins "Professor Popkiss" and "Doctor Beaker"? Not sure about having a monkey in a shiny new car, though.
My coat is the anorak with the Fanderson badge.
If there is a chip, there must be some form of multiplexing going on because it will need some power. I suggest that common ground, left, right, power, and control signal. But there are tricks with variable DC offsets that could be applied, and it is possible to use one line for both power and control signal, but common ground would be easier.
Token Ring vs. Ethernet
Ahh, but with your 16Mb/s token ring installation using mechanical MAU's, woe betide you if someone in the building attached a 4Mb/s device. Instant mahem, with no easy way to isolate the problem, especially if you did not know the order of the desks on the 'ring (the dreaded 'beaconing' error message.)
This took out a major UK IBM support centre for a whole day (we had to rely on 3270 over co-ax using MYTE on the few desks that had PS/2's on them). We then were tasked with splitting the ring in two, with a bridge tying them together to make a single network, so that if the same thing happened again, only half of the desks would loose their network connectivity (and we had to duplicate routers, 3270 gateways, printers, etc.)
Mind you, trying to work out what happened when someone unplugged the terminator of a 10base5 ('thick wire') or 10base2 ('thin wire') Ethernet had it's own problems as well!
And does anybody else remember the 'jabber' light on early 10baseT twisted pair hubs (not switches). Often caused when someone plugged an RS232 terminal into a structured cabling port that was connected to a hub.
I always felt uneasy whenever I had to tap into a 10base5 thick wire cable. It just felt wrong to take what was effectively a drill (the 'tap tool') to the cable that you had spent so much money and time having installed.
Thank god we now have intelligent switches, with automatic isolation of noisy/broken cables, and proper switch mode to eliminate the need for CSMA/CD.
I always believed that the reason why testicles were on the outside was that the optimum temperature to keep stored wrigglers was a few degrees below body heat.
This is normally quoted as the reason why you should bin the Y-fronts or briefs for boxers if you are trying to procreate. Apparently, higher temperatures make male gametes lazy, like Paris on a sunbed, so they don't move as fast.
I once saw a Julia Set (often mis-identified as Mandlebrot) program written in PostScript. Send it to the printer, and wait for hours for it to spit out the page!
To anybody who is still reading.
I'm not going to make another post on this thread after this one.
I don't think that Alexander and I are actually commenting on the same thing. I did not say that I don't use MS software. I said I try to avoid it where I can. But... I currently have 7 systems running XP, 2 running Windows 2000, and a couple running older versions of Windows. Hardly MS free.
This is mainly because they came with Windows. The rest of my family use Windows, except my daughter, who uses MacOS. I am their technical support department, so I get to see lots of Windows problems, with networking, printing, Office software and many other obscure problems, including the normal gamut of viruses and trojans. Linux definitely wins here.
I have not used Vista, I admit, but I would say that in this case 'better' is subjective. I read C very well, and I have looked at Genetic UNIX source (AT&T derived), BSD and Linux. I know much of the philosophy behind UNIX development, having worked inside AT&T and IBM. Generally speaking the code is very, very good in in all cases. I have not seen the Windows code, but from what I have heard and seen, some of the Windows code is not actually understood by Microsoft (one rumor I have heard is that IBM still provide support for some of the OS/2 derived code in the UI). Obviously, this is information by proxy, but what I've heard can't all be wrong.
Of course, Vista is supposed to be a significant re-write, and if the reported resource use is as bad as it sounds (even now), then there is something quite wrong in Vista. Do you think that Windows 7 would be getting as much exposure as it is if Microsoft had not finally recognised that at least the perception of Vista was flawed, even if they do not think the OS itself is.
My personal thoughts about Windows is that the design itself is flawed, in it's security, and it's resource use, and also in the way that users use it. Trying to make Vista secure broke applications left, right and centre, because apps expected to be able to write to strange parts of the filetree.
25+ years ago, Sun came up with a model for using networked computers where a system was never personal. If you used a NFS connected diskless workstation, or even a shared server, your environment moved with you. The systems could all be near identical, and you could log on to whichever one you wanted, and use it as if it was your home workstation. The UNIX security model needed almost no tweaking to make it work, even the split between users and administrators worked. All that was needed was a little segregation of system data into read-only, read-mostly, and read-write data. There was even an application deployment method that allowed you to install the software just on the servers, and have it used on the network workstations. There was even a model that allowed heterogeneous systems in the same environment. This was a sys-admin's dream.
Microsoft in the elapsed 25 years have not managed to come up with a model that works nearly as well. A Windows desktop system is still a Personal Computer in 2009, even with roving profiles, sharepoint, Active Director, and all of the other technology they have rolled out. This is because the basic design is flawed, and there is no point in building on cracked foundations. This makes it basically unsuitable for business, even though much effort has been put in to try to make it so. Just go to a desktop Windows system, log in, and see all of the junk left behind in the copies of all of the profiles of users who have previously used the system. And application deployment? After installing MS Office on every desktop, even with scripted installs, one would have thought that someone would have realized that something could be done better.
And what is being rolled out now? Windows 'Mainframes' accessed via Windows Terminal Server, or Citrix XenApp. Hardly progress. It's almost exactly like IBM's VM/CMS environment (in concept, I'm not suggesting that 3270 terminals ran a GUI).
The world has moved on, and I know that the UNIX NFS model is now dated, particularly the network security. But replace UNIX with Linux, NFS with Kerborised NFS 4, or GPFS, or even CIFS, and the model still works. Linux plugs straight in to this environment, and is bringing in new developments.
It is this type of design that I think is superior. It needs to evolve, and be pushed forward, and I think that groups like X.org are making this happen (the UI has needed a re-work for quite some time), but with the render extensions, and integrated GL in the Xserver, this is happening. Look at Compiz Fusion, and there is scope for UI's to be as pretty and as functional as anything Microsoft or Apple can push out. And guess what. Much of this is being done for free, often by people who code in their working life and contribute on their own time, so can produce good code. Debian is a good distro, and is proved to be so by being selected as the bas for so many other distro's.
This is turning into an essay, so I'll shut up now. To the moderator, sorry I had to put you through all of this. If it is too much, I'll not be too annoyed if you choose not to post it.
@Alexander again again
You have not read my post properly, nor have you answered my challenge. OK, the NHS refused to accept OO documents. You have not said why. Reading between the lines, and from experience, it is probably because although the documents looked OK on the screen in OO, they ended up being formatted poorly in Word. This is a fixable problem that also happens Word-to-Word if you end up with a different printer driver. Tell me if you can say that you have never changed the destination printer in Word, and suddenly found that the layout of the text has changed. If you havn't, then you have been very lucky. Try adding the Microsoft TruType fonts to OO and see whether it is better.
And with regard to training, I think that you have reinforced my argument. I said, and I quote "If your users will only accept MS Office, then OpenOffice will never do". As I said before, I was commenting on home users, and application agnostic users (agnostic in this sense that they are not tied in to an application). Your users were obviously tied.
And again, on the installation. I was proposing having Linux installed, and then explained why it will not happen all the time Microsoft are dominant.
I will admit to being averse to buying Microsoft products myself. And I admit that I do not like Microsoft's anti-competitive practices, as they are morally wrong, and may be illegal. But I do not deny them a position in the market place. Again, I quote from one of my previous posts: "If you want to run Windows, I'll let you. No big."
I question your comments about Microsoft's products being better than their contemporaries. OS/2 was definitely better than Win3.1, and at least as good as NT3.5. DRDos was better than MSDos4 Amipro/Wordpro and Wordperfect were better than Word. Netscape was better than IE. Linux is better than XP/Vista in almost all respects with the exception of application availability.
In almost every case, Microsoft was able to kill their competitors products by means other than technical merit, normally by threats, but also by loss-making pricing by cross-subsidy from other products.
I was not talking SME. I have said this several times, and I say it again. I'm talking SOHO, to use the acronym. SME's have other requirements which definitely make them non-application agnostic.
And I cannot see how ITIL has any bearing on this discussion, either for home users, or even for SME or Bluechip customers. The required procedures with associated documentation when correctly produced will be ITIL compliant regardless of the underlying OS. You could argue that the product documentation needs to be referred to, but Linux has documentation.
And what has six standard deviations from the norm on a distribution curve ("six sigmas") got to do with anything talked about in this comment thread? Do you really know what you are talking about, or are just using buzz words.
As I said. I use Ubuntu because I don't want to spend all my time fiddling. I state that it is possible, because I do it. It's no more difficult that XP.
It is not technical merit that makes system builders decide not to install Linux on new systems. It is the fear that Microsoft will remove them from the OEM list, meaning that they have to buy Windows at list price (£100+) rather than OEM price. This would make the cost of installing Linux very expensive in collateral fallout.
Microsoft can now give XP away, because they have recovered all development costs. They are not competing with themselves, as small systems can't run the bloat of Vista. What they have done, however, is a U turn, because XP was dead in their eyes, and they were trying very hard to make sure it died before UMPC's came along.
And I would clam down if I were you. If your lack of spelling is an indicator, you are stressed.
@Rex Alfie Lee
Don't go on at the community. Rant at the Wireless card manufacturers. If they did their work for Linux as well as they do for Windows (and I don't believe that any version of Windows up to and including XP had ANY wireless drivers at all), then every card would work under Linux.
It is not the Microsoft software that makes the wireless cards work it's that CD that rattles around in the box when you buy it. Most modern Linux distro's will, when installed off the generic install CD, support many of the common wireless chipsets. Not all, I grant, but definitely the Intel Centrino set, the Prism 2 set, and some of the RTL chipsets.
I must admit that I struggled with Ubuntu 8.04 on EeePC 701's. I don't know what they did, but you have to install a modded driver to get it to work. There is a thread on the Asus Community threads that tells you where to get it.
I obviously can't comment on your pharmacy install, but I would expect that the reason why it failed was either the 'it does not look like Word and Excel' argument, or the 'It doesn't run all of the VB code I need' argument.
The first is a matter of training. If your users will only accept MS Office, then OpenOffice will never do. But I would be interested to see how the same users react to MS Office '2007, with it's new look.
The 'it doesn't run all of the VB code' is an example of a user locked into MS products, not the application agnostic users I was talking about. I have this issue in my work, where my work issued laptop has Lotus Symphony (a repackaged OpenOffice), but I need to be able to run a heavily scripted Excel application. So I can understand it, but it is not what I was talking about.
I have other personal experiences of this. My 14 year old son recently was asked to do a PowerPoint presentation for a piece of school work. I suggested he use OpenOffice (installed on pretty much every system in our house), but was told that it had to be done in PowerPoint, because that is what the teacher had demanded. This is an example of the MS monopoly becoming institutionally reinforced, for no good reason. And this is because of the special licensing agreement Microsoft have agreed with most education authorities (== almost free in most cases).
I capitulated, and bought a copy of MS Office Home and Student 2007, which I installed on his gaming rig (see, I run Windows in my home network). It made me feel dirty, as well as a bit bitter about spending money that was not strictly necessary, especially to Microsoft.
But for the average user, buying a system for home use, chances are that they will not buy MS Office anyway. Even at £50 (which is what I paid for Home and Student), this is a significant part of the £300-£500 they pay for the complete system. More likely, they will end up with some variant of MS Works. And even this will provides more function than they will need. I don't see what this can do that cannot be done in OpenOffice, and I use both OpenOffice and MS Office (and also several versions of Works).
Please understand this, I am not, nor have I ever in this comment trail, been talking about corporate or business users. But even if I were, I believe that OpenOffice can hack it as long as there is no other application tie-in. If you disagree that strongly, please give an example of what a user needs that cannot be done in OO. Merely being different does not count, however.
It is interesting you mention Tversity. This appears to be a freeware (for personal use) application, a model that is not that different from that which most Open Source software is published , although a little more restrictive. Yet you have difficulty with Linux and OpenOffice? I detect a degree of hypocrisy here. And it is a stated aim of the Tversity developers that they intend to support MacOS and Linux eventually. See the General FAQ. And for similar functionallity, maybe Elisa may be a better fit on Linux, as it supports DLNA, and is in the Ubuntu repository.
I really don't know what points you are trying to make. If you want to run Windows, I'll let you. No big. But I really feel that Windows is not essential, and a Linux like Ubuntu can currently fit the same space for many or most home users. I know this, because I am doing it. I know exactly how hard it is, and how much of my 1337 Linux haxor skills I have used, and I believe that an ordinary user could do it as well. Can you say the same?
@Alexander and William Henderson.
About the only thing I had to fiddle with for my latest Ubuntu re-install was the DVD codec's. And even that was not difficult, and only not installed initially because it is patent-encumbered. It picked up my wireless card, prompted me for my initial key, picked up address, gateway and nameserver from DHCP, already had installed Firefox that then installed the flash and other bits as required as it would have on Windows, and even discovered the printers in a similar manner to Windows. OpenOffice was configured out-of-the-box. Sound worked, as did USB devices, scanners, and Cameras. Without a single additional CD.
This was significantly easier than the last XP install I did from generic media, where I had to license and activate the OS, and install a whole load of drivers and apps from multiple CD's (including the wireless drivers CD). All with loads of re-boots in the process.
But most users don't do this. They unpack their system from its cardboard boxes, stick it on a desk, and turn it on. This could be achieved by vendors pre-installing ANY Linux distro just as well as they do with Windows. But at the moment, this does not happen. For a brief instant, it looked as if this may have happened in the UMPC and OLPC spaces, but Microsoft crashed those parties by changing their self-imposed rules, and extending the life of an OS they had been desperate to kill a few months earlier.
I state again that I believe that for an application agnostic user (i.e. one who does not need Word, Excel, Internet Explorer, PowerDVD et. al.) but can use OpenOffice, Firefox, and Totem, that Ubuntu installed out-of-the-box will satisfy them, and can be easier. I would state that a gamer does not fall into that category, as they want a particular OS locked application, their games.
For streaming multimedia, flash just works, and Silverlight (and WMP based delivery methods such as those deployed by SKY) fall into the category of artificial barriers raised by the vendors. Only Windows will ever satisfy that, even if Moonlight takes off (BTW, try using the ITV streaming service that even though it installs Moonlight, fails to work). Artificial barriers.
For media serving applications, you should try a packaged MythTV or use the Medibuntu repository, just like you probably would use Windows MediaCentre, rather than using a generic user-targeted Linux distro like Ubuntu or Debian. I don't know whether this would surprise you, but Tivo boxes run Linux under the covers.
OpenOffice does have it's own, internationally agreed document formats. It also understands .doc as well as several other proprietary formats. No prop needed there.
Oh, and by the way, William Henderson, I understand rootkits, and have followed many of the ways they are deployed, and most of them require very specific vectors to get into a system, which probably will not be on an end user's Linux workstation. I know it's possible to compromise a Linux system, but it really is not as easy as a Windows. And the information to fix these is not obscure, just rarely needed. I welcome the day when there is no need for copious tools and web-pages to protect and disinfect Windows systems, when they become unnecessary. They are only out there for Windows because the exploits are there. Supply and Demand.
And anyway, there are Virus and Trojan scanners for Linux. Clam, F-Secure, Kaspersky, Trend and Grisoft all have products you can install/buy, just like Windows.
Linux is not out there in the pre-installed space, because Microsoft threaten vendors who ship systems with Linux installed with withdrawing their OEM status for Microsoft products. This is not hearsay, but a matter of historical record (see recorded cases about Netscape, and Lotus and WordPerfect office suites). This is enough to put a system builder out of business in the current climate, at least until Linux is widely accepted. But this sounds like a circular argument to me.
So Linux is not being judged on merit, it is hamstrung and hampered by those who would lose out if it really became successful. This is why there has to be a unified front from those who know, and why Microsoft should be picked up on all of their anti-competitive practices.
OK, I'm off my soap-box for now, I'll wipe the froth from my mouth, and will pick up my coat as I leave.
Good god, are we really trying to stop Linux!
This comment trail is painful to read, and infuriating.
I am a self confessed geek, and have been for 40+ years, and also a 30 year+ UNIX user (and Linux since it has been around). I've supported UNIX systems at a source code level, and have worked in a major vendor's UNIX support Centre as one of the senior techies.
I have NOT got the spare time to fiddle with Linux to get it working on my day-to-day system. I do NOT want Windows. So, I use a major distro expecting this to do most of the hard stuff, and this has become Ubuntu. It's quick to install, supports pretty much everything on mainstream PC's and provides the necessary apps. for close on 100% of users who are application agnostic (not everyone, I know).
But I can tweak it if I want. It's close enough to Debian to allow me to put most .deb packages on. And I can compile the stuff up if I need it (e.g. the airprime module to speed up 3G USB dongles). If you want to use the bleeding edge Nvidia drivers, you can, but if you don't, what is in the repository will suffice. But I won't deny other distro's the right to exist, or peoples right to use them. Nor will I deny the rights of people to use Windows when they know no better.
It's important for Lenny to be produced, because it feeds through to other distros, but it is not the OS for the masses, and the Debian core team probably know that. Ubuntu could be, but the jury is still out, waiting for mainstream app. and game support. Normal users want stability, confidence that the system works, the apps. they need, and maybe a good update and patching process.
Also, it would not be the first time that a Microsoft product deliberately offered poor support or artificial barriers to alternative technologies. I used to use the Microsoft tools Virtual Desktop (on a work provided laptop, please note) which used to lose windows when switching between desktops. But only the Firefox ones!
To all of the Linux proponents out there, get with this message. We need a DOMINANT Linux distro for the masses. Stop squabbling in public, it's ugly, and makes people turn away.
I agree. The old IBM branded Thinkpads, built in places like Greenock were rock solid.
I pick up my old T23, and still think it is a good piece of kit. My current personal workhorse is a T30 (on which I am typing this), which shows where things started going wrong, but is still quite nice. I have just been provided with a T60 by my employer, which I am beginning to hate. I mean, why put all the sockets on the side rather than the back?
Once Lenovo started adding their own ideas, they lost sight of what business laptops were all about, and just became another manufacturer.
time and reference
I'm sure that the reason why leap seconds are necessary is that the universe is only perfect if treated holisticly.
I don't believe that any scientist or mathematician is arrogant enough to be able to claim that they can take into account all of the significant gravatational objects that could affect both the Earth's orbit or it's rotation. I'm fairly certain that a large asteroid strike, volcanic erruption or earthquake could affect the rotation of the earth, and the solar wind pressure may perturb the orbit enough to introduce a measurable variance in it's orbit. And that's not to mention Andromeda.
So no, the Earth's orbit is not mathematically perfect.
And anyway, who told God that the rotation and orbit of the Earth should follow some nice fixed relationship. It's all a co-incidence.
Is Paris enough of a heavenly body to shake the world?
Open vs. Closed
For goodness sake. It will be something like 10 years before these ships will have another major refit, so will they be stuck with WforW for all of this time? I'm sure the Type 45s are expected to have a 30 year+ lifetime.
With Windows, the MoD are beholden to Microsoft, and have to negotiate extended support at whatever price MS want to charge (how long will Win2K have been out of support when the next refit comes)?
At least with Linux, as a supplier, you can fork the code, and take complete ownership of the product. The length of support becomes how long you are prepared to pay people to be familiar with it, and the security becomes as good as the people you employ. If they are good, you don't even have to rely on the community for fixes. I would have to check, but I believe that provided you keep a product in-house (or should that be onboard-ship), you would not have to release any modified source.
I suppose that one good thing is that if they are using commodity hardware, then it would be possible to drop in the latest Dell server into the rack, provided that you can get the Win2K drivers for the SATA disk and display adapters. Or maybe retrofit another OS.
Pay vs. Free TV
At the risk of starting a major flamewar, I think that some of the people commenting about reducing what they pay for TV should look at who actually pays to have new material created.
There are basically four models for paying for content.
1. Public funding
2. Subscription charging
Please note that there is a significant overlap between the last two (company sponsored programming). And I am aware that there is a divide between the content creators and the carriers.
In the UK, we are fortunate to have a publicly funded broadcaster who has money to commission good content (catagory 1). This means that some of the rating chasing necessary to attract the diminishing advertising revenue is not necessary, and gives the programme makers some leaway. This is what pays for much of the BBC's output, and a significant part of what appears on the free and pay channels Discovery and the UKTV channels carried on Sky, Virgin and Freeview/Freesat.
But the rest is a mismash, mostly catagory 2 and 3. If costs go up (as they do), then either advertising revenue has to increase, or subscription charges go up. Ditto loss of advertising.
But consider this. Let's say that the ratings on cable/satellite go down. The deals with the advertisers are based on numbers of views of the adverts, so advertising revenues go down. So subscription fees must go up (especially if ratings reductions are a result of lost subscribers). This is also what must happen if the subscriptions become more specific, allowing customers to select their exact package. The alternative is that the amount of newly comissioned material goes down, anfd all we get are repeats, which are cheap.
I'm sure the model is subject to tweaking, but if you want a good selection of new programming, which must include subjectivly good and bad material (everyone has their own preferences, like I don't like SpongeBob), then you must expect to pay for material you don't like.
Like may things, it is a compromise, a bit like democracy. Some of you just want something for nothing.
I'm no lawyer, but I believe that GPL3 removes the requirement to publish the source for "system libraries", that include all OS libraries and runtime libraries for languages and tools like Python and Ruby. This is how I interpret the information explained at http://www.gnu.org/licenses/quick-guide-gplv3.html. The test appears to be whether the library could reasonably be expected to be freely available.
In the case of a port to new hardware, this test may fail, and may require all source to be published.
Of course , it is likely that the Linksys products were published under GPL2.
I'm not sure that I understand what the infringement is either.
As I understand it, CISCO can only have infringed the GPL if they have modified any of the GPL'd code, and put it in a product without making the changes they made to the source available. If this is the case, all they need to do is to provide, on request, the altered code to whoever asks for it. They do not even have to put it on a publicly accessible web site.
There is certainly no block on USING, say the GCC compiler, libraries and other tools to create a commercial product which is NOT covered by the GPL, provided that no modified code licensed under the GPL is included in the product. If this were not the case, then any software developed using GPL'd tools would have to be published under the GPL, a situation that is explained at some length in the information about the GNU Public License.
Linking against the libraries is allowed, but modifying the libraries, compiler or tools to produce the product without making the source changes available is not. I would assume that the process of porting to a new architecture which required library or compiler changes would require those changes to be made available.
I would expect that if the libraries were modified as part of the porting process to produce the product, then these changes would have to be published, although it may be a moot point as to whether using a modified compiler counts as included code. It would certainly be against the spirit of the GPL, but it may not break the license in the strictest sense. This one is probably for the lawyers.
Just my two-pennies worth, which btw. can be found in the right pocket.
I'm not typical. I know what I am doing, and I use my eeePC as an adjunct to my existing systems. I got tired of Xandros after the 2nd re-install (I'm sure there is something wrong with the UnionFS implementation on it), and put Ubuntu 8.04 on it. It works fine, and it is just sooo much easier to get out when I need to look something up on Wikipedia than my full sized laptop. I've also used it as a roving WiFi tool when sorting out problems with larger networks.
I even use it to vnc to my other systems, although the small screen size is a bit of a bind. I also had to install a flash-blocker in firefox, because multiple flash adverts can really sap the life out of the processor.
I generally carry around 8GB SD cards with ripped film and music on them. It's amazing how many will fit in your pocket. For this type of content, it is not really a problem swapping them around. I intend to keep using it as long as it works.
I'm sure that the old license contained a phrase about "owning equipment capable of receiving a TV signal", but that predated broadcasts over the Internet.
I got involved in an Email discussion with Castle Communications (who issue and enforce the TV license) about computer based freeview adapters, and was shown the earlier quoted text, but it had not registered that this clause gave you a way out. I found out that you can use mobile (i.e. not mains powered) TV receivers legally anywhere as long as you have a TV license at home, but as soon as you plug it into the wall, it is illegal unless the location is covered by a license.
Remember, it is up to them to prove that you are using TV receiving equipment, and they do not have a statutory-right-of-entry to your house.
Mine has the TV license in the inside pocket, in case I am challenged when using my laptop (please note, ON BATTERY POWER) to watch TV.
Maybe not as good but...
I took delivery some years ago of some IBM kit, in a big box on a pallet.
We opened all of the smaller boxes that we recognized, and found that we were left with one moderately sizable one (about CRT monitor size). We opened it, wondering what extra goodies we had been sent, only to find it was completely empty. Apparently, it was to fill the box to keep the other stuff safe. This was apparently normal practice, and IBM had a range of empty boxes, all with IBM part numbers, to serve this function.
The DVLA is actually an "executive agency", not a government department (the clue is in the name, the Driver and Vehicle Licensing Agency). This makes it responsible to the Department of Transport, but it is actually at arms length from direct ministerial control. The DoT and Parliment make policy and statute, and the management structure of the executive determines how this policy will be carried out (at least in theory).
This makes it interesting in the case of some of the security leaks from executive agencies, as in theory, the management of the agency should resign as being responsible, not the ministers.
I subscribed to Sky HD last year, and was appalled by the LACK of HD material on the Sky channels. Very little of it was actually HD, much was just up-scaled standard definition (is the Simpsons in HD actually better than SD?) And unlike the BBC HD service (which is FTA, even on Sky HD), I pay a £10 a month premium, as well as paying over-the-odds for being an early adopter (HD box £300, installation etc.) So, Sky are not doing the HD stuff out of the goodness of their hearts.
Compare this to the BBC HD output, and you will see that the BBC are actually commissioning much more real HD content than Sky. And the BBC have developed an HD delivery platform in FreeSat (and also HD FreeView), which other people can build the kit for (Sky and Virgin do not actually *make* their kit, it is subcontracted and then branded)
The real HD benefits on Sky HD appear to be for Sport (which I do not want), Movies (not Sky produced), and the HD documentary channels (a good part of which actually show content part financed by the BBC). I am actually thinking of DROPPING the HD subscription, although the number of Sky channels is improving at the moment.
My Sky subscription currently costs me £576 per year, which actually produces about 8 hours a week (estimate) of new Sky material, and the rest is access to many other channels not produced by Sky (much carrying BBC material - look who owns the UK* channels). My, the BBC license fee actually looks really good value for money, especially now that FreeSat is available as well as FreeView.
The reason Sky want the BBC shutdown is so that they will have a near-monopoly of TV delivery (also the reason why they are so antagonistic towards Virgin buying into ITV) so that any content worth watching has to be paid for, and to BskyB. It's all down to the money and profit.
I agree that the license fee is effectively a tax on owning TV receiving equipment, but why not just think of it as a subscription to whichever part of the BBC you use. It's still good value for money, even if you only use the radio, or news, or the FreeSat or FreeView delivery platform.
Xandros not up to scratch
Not sure what happened, but on my 701, I managed to fill the root filesystem. but no way could I free up space. Somehow, whatever I wrote on the root fs got copied onto the unionfs copy. Got to the point that the system no longer booted. Tried booting from a USB flash drive (which worked), but was just unable to free any space at all.
This was the second time I had Xandros screw up, and I am a long-term (10 years) Linux power user (and even longer UNIX), and could not fix it. Xandros appears to have a really strange startup process that I just could not get to grips with. Decided to put Hardy onto the system, and once the drivers were sorted, system works fine.
I like the idea of unionfs, but I just could not fathom what was going on. As far as I was aware, the read-only copy was supposed to never change, so you could restore the system to just-installed condition.
Xandros is probably fine for the simple mode, but I would not choose it as a general purpose Linux distro.
With regard to Vista
OK, we all know that Vista Premium or above works OK when installed from scratch on a contemporary PC. I would hope that if a hardware vendor expecting to sell a system with Vista would spec. the hardware out properly to give a good user experience.
Where I have a problem is them trying very hard to quash support for the perfectly usable 3-4 year old systems that still run XP very well, but do not have the oomph for Vista. Are Microsoft in league with the hardware manufacturers?
I tend to not dump systems that still work, so have three systems running XP on AMD and Intel processors between 1 and 2 GHz and using 256-512MB of memory and modest AGP graphics cards. The kids use them for their homework. So, I buy a new HP printer, and find that I cannot load the drivers on an XP system anymore, even though the printer used to have an XP driver disk (I actually have two of the same printer). The reason? Microsoft have forced other vendors to withdraw drivers that were designed for XP using clauses in their Windows licensing agreement. HP even said as much. They do not carry the drivers on their support website, and provided a helpful sheet of paper with the printer saying that I should keep the XP driver disk safe (the one with my earlier purchase) as HP would not be able to provide the driver after a certain date.
Not sure how the climbdown over XP sales for UMPCs have affected this.
Also, how long are other software vendors going to be able to produce AV, firewall, and other required pieces of software that rely on XP libraries and DLLs. I'm sure that when Microsoft actually decide to kill off XP for good, they will attempt to make all necessary utility software writers to dump XP as well using similar clauses.
I understand that Microsoft may not want to offer full support for older OS releases, as they have a business to run, and providing support forever is just not profitable (but I bet the US military get 10 years support from product withdrawal). But to try to get other businesses to drop support just to make users replace usable computers is just wrong.
Generally, my kids want Windows (one of my sons was very upset the other week because I do not have MS Office to allow him to continue some school work at home), even though I am a committed Linux advocate. Microsoft are just using their dominant position unfairly where ever they can. I know I can get a student edition of Office, but that is still £85 or so. Open Office is free, so why don't schools use it? Because Schools can buy one copy of Office, and then use it on as many systems as they want (and teacher's own PCs as well) without extra payment, effectively free. What chance have other software vendors got?
Anyway, rant off.
Free OS does not prevent paid-for software
Just because the OS is free, this does not mean that a software house cannot charge for its wares that run on that OS.
The various public licenses have clauses that state that you cannot incorporate free code in charged for software, but they normally also allow you to compile against libraries from systems, and also to use the command sets in software.
So, if you have an idea which can result in a software package, it is perfectly possible to charge for that software. You don't have to contribute it back to the Open Community unless you have taken GPL or similar code and incorporated it into your package (and some of the licenses also allow a degree of that). But you can use GCC, Perl, PHP etc. to create the software including the libraries. I have wanted a HMRC certified small business payroll for Linux for ages, and would be prepared to pay the same for a Linux package as I would for a Windows package.
It is only where there is an already available good free package that you would have difficulty in selling your software. If you can sell a good DVD creation or audio editing package on Windows, you can do the same on Linux. Where is the difference? It is only people who do not understand the licensing model who think everything on Linux has to be provided free. Has the availability of Audacity on Windows stopped people selling audio editors on Windows? No,
I would agree that the proliferation of packaging tools is a problem, but you should really treat each distro as a separate OS, at least until a common packaging tool is agreed on. Until that time I will continue making the suggestion that Ubuntu is as good a candidate for the dominant distro as any.
@AC re. NT and USB
I think that you will find that NT predates Win95 by some years (even NT4 was about the same time as Win95, so predates OSR2).
Do you remember USB on Win95OSR2? Yes USB was in the OS, but you had to load 'drivers' for each device (it did not understand device families, so needed the USB ID for the device to be added), so plugging in a new memory stick or printer required you to put the driver disk in before you could use it.
I added USB support using vendor supplied OS extensions to NT4 on both Compaq and Dell PCs. Was not complex, and worked at least as well as in Win95OSR2.
@AC - WTF?
In case you had not noticed, consolidation in the services and hardware business has been going on for the last 30 years. Ever wondered where Tandem, Compaq, Digital Equipment Corp., Pyramid, Data General, ICL, Amdahl, Sequent, Nixdorf, NCR, Honeywell, MIPS (I could go on, the hardware market only really has three or four major players in the non-PC space). All of these have been subsumed by larger companies.
If you think that this has made it easier for startups, then you have a distorted view of the hardware market. Study what has happened to Transmeta or Inmos, who were new companies built around innovative products.
What now happens is that a good startup produces a good product, fails to get capitalized to exploit the product properly, and promptly gets bought by the larger players.
Just wait for IBM, AT&T, and Microsoft to start stifling new startups by leveraging their patent portfolio. It will be nearly impossible for innovators to even get started.
All that the financial mayhem will do is shake the market out even more.
The same will happen in the software market. Just how many companies have been bought by Microsoft, Oracle, IBM, SAGE, Computer Associates (spit), Google, Symantec et. al. because they are competitors or have products that are genuinely new.
I agree that nobody is producing good software for SMEs at the moment, but that will not change, it will just get more difficult to start. There are just too few people prepared to venture money at the moment.
I will state an interest here. I have been an IBMer for seven years (although I missed the blue blood transfusion - I kept to my UNIX roots), and currently get a lot of my work on IBM kit. IBM are not perfect, their software is patchy, rushed to market, and the quality has gone down in recent years, but you are not going to get government agencies, Blue chip companies or major utilities buying software from a startup. Their buying policy will not allow purchases unless companies have a history and a good credit rating. Sad, but true.
All that rationalization will do is to remove choice that will not been replaced. Besides, IBM is reporting GOOD figures (growth is growth), and is extremely unlikely to go under, as they are not exposed to the credit market, being cash rich! Far more likely is that Sun will go. IBM or Microsoft could buy Sun out with the change in their pockets, especially if their stock crashes.
I would hate the hardware market to be IBM, HP and the PC manufacturers. We need competition, and currently only large players can compete.
@Horse loving AC
You miss out the rest-stops, maintenance (vet and farrier) bills, expensive garaging, requirement for fuel even when not in use, and basic stubborness whenever its advanced autopilot decides that it does not like fluttering plastic bags, flashing lights or even drain covers.
Also, are you going to invent the waste collection services, because I believe that before the advent of the car, major cities would be inundated with piles of horse sh*t all over the place.
- Product Round-up Smartwatch face off: Pebble, MetaWatch and new hi-tech timepieces
- Geek's Guide to Britain The bunker at the end of the world - in Essex
- FLABBER-JASTED: It's 'jif', NOT '.gif', says man who should know
- If you've bought DRM'd film files from Acetrax, here's the bad news
- VIDEO Herschel Space Observatory spots galaxies merging